The Score is IBM - 700,000 / SCO - 326

The Peanut Gallery writes "After years of litigation to discover what, exactly, SCO was suing about, IBM has finally discovered that SCO's 'mountain of code' is only 326 scattered lines. Worse, most of what is allegedly infringing are comments and simple header files (like errno.h). These probably aren't copyrightable for being unoriginal and dictated by externalities and aren't owned by SCO in any event. Above and beyond that, IBM has at least five separate licenses for these elements, including the GPL, even if SCO actually owned those lines of code. In contrast IBM is able to point out 700,000 lines of code, which they have properly registered copyrights for, which SCO is infringing upon if the Court rules that it repudiated the GPL."

Linus says he wrote errno.h himself

[stratjakt]

At least the wikipedia article says so.

Is he lying or not? If the original unix comments are in there verbatim, it sounds unlikely that it was completely original.

I'm not saying it should be copyrightable or affect the suit at all, but it certainly bears on Linus' credibility, if he copy/pasted a header file then claimed it was a product of his genious.

Re:Linus says he wrote errno.h himself

[stratjakt]

No, I could see him writing "#DEFINE FALSE 0, #DEFINE TRUE !(FALSE)" and believe it.

But comments are written in human language, and it's unlikely that two people phrase a complex thought th same way. If you're grading programming assignments in university, and see the exact same comments in two student's works - it pretty much tips you off to cheating.

I don't know what Linus actually said - whether he copied from Minix, the wiki article says "Linus Torvalds, the creator and trademark holder of Linux, has denied SCO's claim, saying he wrote the code himself."

If he copied it from Minix, he didn't write it himself.

I'm not talking about IBM, SCO, or anything else. Is Linus a liar, or is the wikipedia article misrepresenting what he said?

Re:Just 1 function.....

[bl8n8r]

> It doesn't matter if it is only 1 function that IBM has copied in there.

RTFA. They are talking about function prototypes, not functions. Big difference. Without actually seeing what the beef is, SCO's claims could be as ridiculous as "int foo (void);"

> Just the fact that SCO has not been lying is vindication enough for me.

Where does it say SCO has not been lying? RBC, Microsoft, SCO and Baystar capital* have been in on this pump-n-dump since day one. As far as I'm concerned, they are all crooks and should be brought to court and tried as such. It's no different than Enron and the other MegaCorp swindlers.

[*] http://news.com.com/Fact+and+fiction+in+the+Micros oft-SCO+relationship+-+page+2/2100-7344_3-5450515- 2.html [com.com]
http://www.newsforge.com/comments.pl?cid=87796&sid =36545 [newsforge.com]

Re:SCO stock

[badasscat]

100 lines, 12000000 lines, 10 lines, SCO wasn't flat out lying. They found something. That something might be enough to win the suit (maybe not a billion).

They only win the suit if they can somehow convince the judge that none of IBM's licenses apply, including the GPL. And if they convince the judge that the GPL doesn't apply, then they are now liable for the 700,000 lines of IBM code that SCO has appropriated.

So, no, they can't win.

Alphabetic order...

[mengel]

As the IBM guy points out in the hearing (online over at Groklaw, of course) the error values are in alphabetic order with increasing integer values.

Exactly what most people would do in building such a list of #defines...

Re:SCO stock

[nege]

I wonder how much sco.com will be sold for. Somebody should make it point to kernel.org, just as a lesson to the patent trolls.

The internet version of severed head on a pike. I like it!

Re:SCO stock

[KokorHekkus]

...It's been under a dollar a share for a few days now. If it continues, it could lead to delisting. Look for SCOX if you want to track it on a ticker. However, in the past SCO did a reverse split and they could always do another one and convert 2 or 3 shares to 1 and get back over a dollar a share to avoid delisting... ...I'm hoping for a delisting as that would hurt SCO immensely, but I'm not holding my breath

As you suspect there isn't much hope for delisting. The stock has to be under $1 for 30 trading days for the company to get a warning. They then get 90 days to remedy the situation (for example, doing a 2-1 reverse split on the stock you suggested). It would be harder for SCOX to avoid a delisting if they slid under the other delisting criteria: having a market capitalization under $10 million (currently it is about $20.4 million). But as you, I'm not holding my breath for either one.

Re:Linus says he wrote errno.h himself

[sconeu]

Linus reaffirmed that in a story on Groklaw [groklaw.net].

Waaay OT, but...

[Spaceman40]

...isn't it interesting the lines that are most common?

I can tell that you're coding in C++ because of the private/public and the }/}; (that inconsistency has always bothered me: is it a statement or not?).

I ran this* on the Python files of the Django [djangoproject.com] project, and got some interesting results:

11185:
2314: """
1205: else:
1063: try:
288: pass
269: ...
235: Traceback (most recent call last):
226: from django.conf import settings
185: }
164: )
148: def __str__(self):

Interesting comments:

• Yes, that's right, 11,185 empty lines. We programmers sure like our whitespace.
• """ is the delimiter for Python-friendly documentation. (Think javadocs.) If they documented each method and class, you could grep for def and class and get an estimate of how documented the project is.
• pass basically marks an abstract/unimplemented method. I'm running straight from Subversion, so that may have something to do with it.
• While Python doesn't use {} for methods, it does use them for dictionaries. The convention for multi-line dictionaries is to have the closing bracket on its own line.
• __str__(self) is Python's analog to Java's toString or Ruby's to_s.

Interesting stuff. I wonder what it would look like in Haskell, or Lisp...

* I wrote a modified version in Python that walked the directory tree and stripped lines of whitespace, otherwise it was pretty much the same: Simple statistics. [dpaste.com] Yes, I know it's somewhat messy. I tried to clean it up a bit before putting it up. A few simple modifications would make it work with any extension.

Re:SCO stock

[TheViffer]

Wonder how deep Microsoft's wallet will go on this ...